paraffin

Subscriber access provided by TUFTS UNIV

Review

A perspective of solutions for membrane instabilities in olefin/paraffin separations: A review Antoniel Carlos Carolino Campos, Rodrigo Azevedo dos Reis, Alfredo Ortiz, Daniel Gorri, and Inmaculada Ortiz Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.8b02013 • Publication Date (Web): 09 Jul 2018 Downloaded from http://pubs.acs.org on July 12, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

A perspective of solutions for membrane instabilities in olefin/paraffin separations: A review Antoniel Carlos C. Camposa,b, Rodrigo A. dos Reisb, Alfredo Ortiza, Daniel Gorria, Inmaculada Ortiza* a

Department of Chemical & Biomolecular Engineering, University of Cantabria, Av. Los Castros

s/n., 39005 Santander, Spain. b

Institute of Chemistry, Rio de Janeiro State University (UERJ), Campus Maracanã, P H L C,

São Francisco Xavier St., 524, Rio de Janeiro, RJ, Brazil, 20550-900. *

e-mail: [email protected]

Submitted to Industrial and Engineering Chemistry Research Revised manuscript – July 2018

ACS Paragon Plus Environment

1

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 59

ABSTRACT:

Light olefins are mainly produced by the naphtha steam cracking, which is among the more energy intensive processes in the petrochemical industry. To save energy, some alternatives have been proposed to partially replace or combine with cryogenic distillation, the conventional technology to separate olefins and paraffins. Within this aim, facilitated transport membranes, mainly with Ag+ cations as selective carriers, have received great attention owing to the high selectivity and permeance provided. However, to be used industrially, the undesirable instability associated to the Ag+ cation should be considered. Poisonous agents and polymer membrane materials are sources of Ag+ deactivation. In recent years, great achievements on the separation performance have been reported, but the current challenge is to maintain the selectivity in longterm separation processes. This work presents a critical analysis of the potential causes of Ag+ deactivation and points out some alternatives that have been proposed to overcome the hurdle. This review highlights and critically analyses some perspectives of the ongoing development and application of facilitated transport membranes.

KEYWORDS: Olefin/paraffin separations; Facilitated transport membranes; Silver salts; Silver nanoparticles; Carrier poisoning.


2

Page 3 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. INTRODUCTION Light olefins are the principal raw material to the petrochemical industry.1 In 2016, the global ethylene and propylene production was 146 and 99 million tons and the expected of demand growth rate, until 2025, is 3.6%/year and 4.0%/year, respectively.2 Steam cracking (SC) is the main industrial process to manufacture light olefins from naphtha or light alkanes.2–5 In petrochemical industry, SC is among the more energy-intensive processes, and, in 2016, the worldwide consumption have been estimated to be about 3.0 · 1015 BTU.6,7 After pyrolysis, separation section is the second large energy consumption step in SC (about 7.5 · 1014 BTU).2,5–7 The justification for this consumption is based on the cryogenic distillation of the cracked gases.6,8,9 Membrane processes have been proposed to save energy in the separation section by replacing or integrating with the current cryogenic distillation technology.10–17Membrane technology has achieved very promising results, specifically regarding the olefin/paraffin separations.18–24 In a recent study, Lee et al.25 identified the optimum membrane performance required to replace one typical C3 splitter. They found that a set of membrane modules with propylene permeance of 11.3 GPU (1 GPU=1×10−6 cm3 (STP)/cm2 s cmHg) and selectivity of 68 could substitute a typical distillation process. Since the replacement of the distillation column by membrane units with optimum performance is technically viable, the great challenge for membrane technology is to reach the suitable selectivity and permeability for the process and keep them over long-term operation. The operational conditions in which the membrane should be used are severe and can lead to the performance loss along the operation.23,24,26–28 Besides the steam cracking (SC) plants, which constitute the main target application, there are other two potential areas where membrane technology could be applied for olefin/paraffin


3


Page 4 of 59

separations. Olefin/paraffin separation membranes could be used in the recovery of propylene from FCC off-gas streams29–32 and in vent streams of some kinds of petrochemical reactors, e.g., polypropylene reactors.33 The recovery of olefins from vent streams is very attractive because milder operational conditions are present (low level of sulfur and acetylene compounds) and the performance required is lower (selectivities of 3-5) than needed in SC process and FCC off-gas streams.33 Maybe, this market niche can represent a previous favorable step before the attempt to replace distillation in the SC process or in the FCC off-gas streams that involves huge challenges to be overcome.31,33 Among membrane technologies, the alternatives based only on the solution-diffusion mechanism are not effective enough to discriminate olefin/paraffin pairs. The similarity between the physico-chemical properties of alkenes and alkanes is the main drawback that all dense-type membranes assigned to the separation face.34 The carbon molecular sieve membranes suffer of the same problem, since there is a tiny difference between the molecular diameter of the molecules to be separated.15 Nevertheless, many efforts have been focused to improve the separation using mixed matrix membranes by introduction of zeolites,35,36 organic and metal−organic frameworks37–41 in the polymer matrix. Other works focused on carbon molecular sieve membranes10,42,43 that are brittle and difficult to scale-up the production.25 To overcome these discriminative issues between the molecules, the facilitated transport of olefins has been explored, increasing simultaneously the permeance and the selectivity of the separation. The facilitated transport can be defined as a process in which chemically distinct carrier species form complexes with a specific component in the feed stream, thereby increasing the flux of this component relative to other components.44 The facilitated transport membranes have surpassed


4

Page 5 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the upper bound of the Robeson diagram for the olefin/paraffin separation in the last few years,22,45 demonstrating the greater potential of this technological option. Considering the potential applicability of the membrane technology for light olefins/paraffins separation, the goal of this work is to critically review the development of facilitated transport membrane for this separation highlighting the challenges and main drawbacks surpassed during the last decades. A comprehensive analysis of the instability/deactivation problems confronted by distinct kinds of membranes is carried out. The alkene/alkane membrane separation technology is presented describing the source of poisonous agents for the principal carrier used, i.e. the Ag+. Finally, some recent strategies are pointed out as options that try to overcome the Ag+ deactivation by smart solutions.

2.

FACILITATED

TRANSPORT

MEMBRANES

FOR

OLEFIN/PARAFFIN

SEPARATION The development of new membranes based on polymeric materials or the modification of polymer structures that provide suitable selectivities for the olefin/paraffin separation has been largely investigated.46–48 As the polymer films have become more selective to the olefins, they lose productivity due to the trade-off between selectivity and permeation flux, therefore, leading to the technical unfeasibility of the separation process.15,49 Films where mass transport is based only on the solution-diffusion mechanism have shown, as best results, ideal selectivity of 27 (C3H6/C3H8) and 0.8 Barrer of propylene permeability.48,49 These values are significantly lower than the performance reported for facilitated transport films that have shown mixture selectivities (C2H4/C2H6) higher than 100.45


5


Page 6 of 59

The facilitated transport mechanism, when it takes place, plays in a complementary mechanism to the solution-diffusion transport. In the facilitated transport, a carrier agent interacts reversibly and selectively with the target molecule, that results in the increase in the driving force for the permeation flux of target molecule, and therefore, the permeate is enriched. This kind of transport is not accessible to the inert molecules that are not able to interact with the carrier agent, so that their concentration decreases in the permeate side (Figure 1). Thus, the facilitated transport membrane has a superior selectivity compared to the passive (solution-diffusion) membrane.27,45,50,51 The carriers need to be effectively dispersed over the natural diffusional path of the gases (in the direction of the concentration gradient) and be present at a concentration high enough for transport activation. Furthermore, the carriers also need to be ready to the interaction with the target molecules inside the membrane, i.e., the carriers should be not poisoned.52,53

Figure 1. Representation of the facilitated transport mechanism through selective membranes.


6

Page 7 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.1. Chemical interaction for olefin separation: Olefin π complexation. The formation of complexes between some metals and the double bond of olefins has been well known since longtime. Nevertheless, only in 1827, the first metal-olefin complex was identified. The referred compound was the platinum (II) -ethene, which was known as Zeise complex. In the beginning of the 20th century, the first ideas of using silver (I) (Ag+) salts in olefin absorption systems arose. However, only in 1951, Dewar54 gave a satisfactory explanation to the interaction mechanism between the ethene and Ag+. Shortly after, Chatt and Duncanson55 advanced the Dewar`s explanation presenting the interaction mechanism called π-bond complexation.13 This complexation takes place when the bonding orbital of the olefin donates electronic density to the empty outermost orbital of Ag+ (5s) making a σ bond. The strength of this bond depends on the magnitude of the metal positive charge (e.g. silver, copper, and gold). The second bond formed is a π bond, resulting from the backdonation of the electronic density from the outermost atomic orbital 4d, which is electronically completed, to the π*- antibonding molecular orbital of the olefin (Figure 2).56,57


7


Page 8 of 59

Figure 2. π complexation between olefin and Ag+ ions (Adapted with permission from Eldridge.58 Copyright 1993 American Chemical Society).

In 1960, Scholander59 published a pioneering work that explored the facilitated transport in liquid membranes containing hemoglobin as carrier agent for oxygen transport. Thereafter, works that described membranes of facilitated transport have grown for several applications.60 Over the following decades to the present day, the facilitated transport mechanism has attracted great attention of many researchers due to the separation potential compared to the simple mechanism of passive transport. Practically, the development of facilitated transport membranes for light olefin/paraffin separations has been based on the feature of the reversible interaction among olefins and some transition metals, especially silver, copper and gold. Additionally, the π complexation should be strong enough to favor the interaction between the metal and at the same time allow the complexation reversibility under the appropriate operational conditions. Among transition metals, silver has one singularity related to π complexation. The silver electronegativity is 2.2, in the range 1.6 - 2.3 in which the reversible complexation can take place.61 In addition, the silver salts that have been applied in facilitated transport of olefins have the lowest lattice energies compared to other metallic salts that can be also used for this goal. A salt with low lattice energy favors the solubility of the metal cation and hence it also favors its availability for the interaction with the olefin.27,62 Owing to these features, silver is the metal most widely used in the preparation of facilitated transport membranes for olefin/paraffin separation. Nevertheless, the use of copper is the second option due to the lower price of this metal25. The price is a critical issue when the final goal is to scale up the production of this kind of membrane and therefore it


8

Page 9 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


should be considered. However, up to now, the use of the Cu+ cation has not proved to be a feasible option for the facilitated transport of olefins due to the stability issues of this cation against oxidizing agents,63 although some promising studies have recently been published about the use of ionic liquids as solvents that stabilize cuprous ions.64 Based mainly on the use of silver salts, the principal carrier, facilitated transport membranes for olefin/paraffin separations started to be developed initially as supported liquid membranes. After, in the search for superior mechanical stability, ion exchange and electrolyte membranes were applied to the olefin/paraffin separations.

2.2. Facilitated transport membranes. 2.2.1. Supported liquid membranes. In 1973, Steigelmann and Hughes65, working in the Standard Oil Company, started to develop films of cellulose acetate with silver nitrate solution in the pores of the membrane (support) (Figure 3). The solution is held in the pores of the support by capillary forces. The best initial result achieved for mixture selectivity (C2H4/C2H6) was ca 1280 and a mixture permeance of 30 GPU. Motivated by the preliminary results, they have developed these films for more than 10 years.66 However, spite of all efforts, they have not had success in the commercialization of this technology. The main problem found by them was the poor stability of the Ag+ solution in the membrane pores. During the separation process, the solution was gradually swept out from the pores due to the dragging effect of the gas stream, dropping the selectivity of the process. To solve this problem, some subsequent works27,67–70 have focused on improving the stability of the solutions inside the pores of the membranes. However, the inherent instability of the supported


9


Page 10 of 59

liquid membranes continues to be a limiting feature in the commercial application of this membrane configuration. Common performances of supported liquid membranes show selectivity values (α) ranging from 100 (α of C3H6/C3H8 using triethylene glycol/AgBF4 43 wt.% with humidified feed stream)67 to 1000 (α of C2H4/C2H6 using AgNO3 4 M with humidified feed stream)71 and permeances from 4 to 27 GPU of olefin, respectively. High permeances are achieved due to low mass transport resistance through liquid medium. The use of humidified feed stream is required to avoid the drying of solution held in support pores. The support can be prepared of microporous membranes made of cellulose, polyvinilydene diflouride (PVDF), and polytetrafluoroethylene (PTFE).67,71–73 Indeed, the disadvantage for this membrane configuration are the real risks of dragging out the carrier solution from support pores.

Figure 3. Supported liquid membrane.


10

Page 11 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.2.2. Ion exchange membranes. In the 1980`s, to solve the problems with the supported liquid membranes resulting from the Ag+ solution sweeping out by the passage of the gas stream, LeBlanc and co-workes74 proposed the use of ion-exchange membranes. From this work, several other groups started to research this kind of strategy.18,75–81 The ion-exchange membranes are generally prepared by the addition of the silver salt to the membrane, which is formed by a polymer functionalized with an acid group, (e.g. sulfonic acid group) able to exchange H+ for the Ag+. To achieve the ion-exchange, the polymer should be immersed in the Ag+ solution or other metallic salt solution. Next, the membrane should be humidified. Without water, the Ag+ ions are so strongly attached to the anionic sites in the membrane that makes very difficult the interaction with the olefin (Figure 4). Working with humidified feed streams, several interesting works have been reported in literature. For instance, Eriksen et al.75 applied Nafion (N-117), which was preswollen in glycerine and soaked in aqueous AgBF4 6 M, for separation of C2H4/C2H6 (1:1 molar ratio) humidified stream. The membrane provided a selectivity of 1930 and C2H4 permeability of 26800 Barrer or about 83 GPU. As the carrier agent cannot be easily swept out from the membrane by the gas streams, the ionexchange membranes have a vast advantage compared to supported liquid membranes. Despite their advantages, ion-exchange membranes formed by an ion-exchange polymer are usually more expensive and the required humidification is not desirable because it requires an additional operation step aimed at drying the outlet gas streams from the membrane unit.27,28 For this kind of membrane configuration, olefin/paraffin selectivity values between 290 and 1900, and olefin permeances ranging from 5 GPU to 83 GPU have been reported (always with humidified feed


11


Page 12 of 59

streams). The main polymeric materials used as matrix for ion-exchange membranes are Nafion and sulfonated polyphenyleneoxide.74,75

Figure 4. Ion-exchange membranes.

2.2.3. Electrolyte membranes. During the 1990`s and 2000`s, to overcome the problems originated from the humidification dependence of the ion-exchange membranes, the development of dense materials denominated as polymer electrolyte membranes took place.20,82–84 Among others, the research groups of Ingo Pinnau83,85–87 and Yong Soo Kang84,88–91 stood out during the last few years. Pinnau and Toy83 reported that it was possible to dissolve silver salts in a hydrophilic polymer with polar functional groups able to coordinate with Ag+ ion, e.g. polyether. In this kind of membrane, the facilitated transport was developed without humidification of the flowing gas, which represented


12

Page 13 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a great achievement compared to the ion-exchange membranes. Kang et al.45,51,62,84,92,93 followed this method and dissolved silver salts in others polar polymer matrices as poly(2-ethyl-2oxazoline) (POZ), polyvinylpyrrolidone (PVP) and poly(styrene-b-butadiene-b-styrene) (SBS). The results obtained with polymer electrolyte materials have by far exceeded the previous results of supported liquid and ion-exchange membranes. In 2006, the Kang`s group,45 using the polymer electrolyte membranes with different silver salts and polymers, was able to surpass the upper bound of Robeson diagram with selectivities and permeabilities never achieved before. For instance, the ideal separation factor (pure gas permeation) of propylene/propane was above 10,000 with 45 GPU of propylene permeance.62 In the permeation of gas mixtures, the selectivity dropped to 40–60 due to the plasticization effect that occurred in the membranes.51 Inside the polymer matrix, the Ag+ cations can be arranged as free ions, ion pairs or higher order aggregates.94 In this context, the term “free ions” should be understood as the Ag+ cations dissolved in the polymer matrix. The best way for the salt to be in the membrane is in the form of free ion, because the silver is more available to the interaction with the olefin.90,94–96 To reach the desired amount of free ions in the electrolyte membrane, normally, the polymer should have suitable functional groups (e.g. ether, amide, lactam, ester, alcohol and aromatic or aliphatic double bond) to interact with the Ag+ cations (Figure 5a). Polymers like poly(2-ethyl-2oxazoline) (POZ), poly(ethylene oxide) (PEO), polyvinylpyrrolidone (PVP), polymethacrylates (PMA), poly(vinyl alcohol) (PVA), poly(styrene-b-butadiene-b-styrene) (SBS), poly(ethylene phthalate) (PEP),45,51,62,84,92,93 poliurethanes (PU) based on polyether or polyester21,97–99 and poly(ether-block-amide) (best known under the trademark Pebax®)28,100,101 are used as suitable polymer matrix to maintain the dissolution of the silver salts in solid electrolyte membranes. The lower lattice energy of the salt is also crucial in this point to provide an easier way to dissolve the


13


Page 14 of 59

compound. Regarding this aspect, AgBF4 is the most widely used salt due to its lowest lattice energy among ordinary silver salts for this purpose.27,62 At this point, in an attempt to increase the degree of salt dissociation inside the polymer electrolytes, a couple of investigations have proposed the addition of asparagine102 in the POZ/AgBF4 membranes and the use of a mixture of silver salts103,104 to improve the salt dissociation inside the polymer matrix. Nevertheless, the interaction among the polymer functional group and the Ag+ salt may cause the reduction of Ag+ cation to Ag0 metallic. Over the time, the Ag0 growth and agglomeration cause the formation of some defects or holes at the interface between the metal particle and the polymer chains. Without discrimination, the gases can easily pass through this path with lower mass transport resistance that leads to selectivity loss in long-term permeation experiments.26 Trying to solve this problem, several works have investigated solutions to overcome this drawback. In 2001, Jose et al.105 retarded the formation of Ag0 by incorporating phthalates to the membranes of PVP/AgBF4 (Table 1). The stabilization of Ag+ cation is due to the strong interaction between the carbonyl groups of phthalates and the Ag+ that plays a key factor in slowing down the reduction induced by the lactam group of PVP. This was the pioneer work that started to report long-term experiments regarding the stability of polymer membranes containing silver salts. In attempt to avoid the Ag0 growth and agglomeration, Park et al.106 added a nonionic surfactant (n-octyl β-D-glucopyranoside (8G1)) to the same kind of membrane to provide a steric hindrance effect hampering the metal particles coalescence. The protective layer onto the surface of formed silver particles was responsible to maintain the stability of membrane performance for 30 days (Table 1). However, the reduction problems were not solved by this strategy.


14

Page 15 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The reduction of Ag+ by polymers like POZ normally results in an increase of H+ ion concentration in the medium.107 This is possible because the membrane contained a small amount of water favored by the hygroscopicity of salts like AgBF4.108 To suppress the Ag+ reduction process, Kim et al.89 proposed the introduction of HBF4 in POZ/AgBF4 membranes. The goal was to shift the equilibrium of the reduction reaction toward the regeneration of Ag+, preventing the formation of metallic silver. To investigate this proposal, they performed permeation tests under UV irradiation. As a result, it was found that tiny amounts of HBF4 could indeed suppress the reduction of Ag+. A POZ/AgBF4 membrane with the molar ratio of 1[carbonyl oxygen]:1[Ag+] exhibited a selectivity of about 100 (50:50 vol.% of C2H4/C2H6) but, after 4 h under UV irradiation, the selectivity dropped to 1. When HBF4 was introduced in the membrane with the molar ration of 1[carbonyl oxygen]:1[Ag+]:0.2[HBF4], the selectivity was maintained in the same initial value after 4 h under UV irradiation, thus having suppressed the Ag+ reduction process inside the material. Although this procedure has been effective in laboratory studies, it seems an alternative difficult to implement on a large scale. On the other hand, Kang et al.109 interestingly suggested the introduction of Al(NO3)3 in POZ/AgBF4 membranes (Table 1) to suppress Ag+ ion reduction. The function of Al(NO3)3 is to weaken the interaction between the functional group of the polymer and the Ag+ due to the favorable electrostatic interaction between Ag+ and NO3-.110,111 The mutual interaction between the ions, i.e. Ag+/NO3- and Al3+/BF4-, is responsible for changing the chemical environment of the Ag+. Compared to the neat POZ/AgBF4, the presence of Al(NO3)3 decreases the binding energy of the valence electron in the silver atom, which is verified by X-ray photoelectron spectroscopy (XPS) analysis. By modifying the electronic density of the silver atom, it is possible to adjust the intensity of the interaction between the Ag+ and polymer functional group,


15


Page 16 of 59

reducing the oxidative power of the Ag+ (Figure 5b). Using this strategy, it was possible to maintain the selectivity of the membrane for 14 days in long-term permeation tests.109

a)

b)

C

C

3-

NO3

O

O Ag+ ||||||| BF4

3+ Ag+ |||||||||||||| BF4 |||||||||||||| Al

-

O C

C

O

POZ chain

Figure 5. a) Interaction between the functional groups of POZ (amide C=O) with the Ag+ from AgBF4 and b) the mutual interaction between the Ag+/NO3- and Al3+/BF4- that weakens the former interaction between the C=O group of polymer.

Also in an attempt to solve the problems related to the essential instability of Ag+ inside polar polymer matrices, Kang et al.90 showed a way to disperse silver salts in polydimethylsiloxane (PDMS), which is an inert polymer matrix, and yet reach the facilitated transport (Table 1). The Ag+ does not share the same interaction observed in polar matrixes, leading to the formation of silver salt aggregates trapped in the polymer domains. At first glance, it seemed that it would not work, since the preferable interaction with the olefin takes place with silver free ions. However, in permeation tests, when the olefinic gases began to pass through the membrane, the olefin complexation gradually dissolved the silver salt aggregates into free ions, promoting the facilitated transport. The time to reach the dissolution was about 100 h, after that, steady-state transport was achieved. Following this approach, it was possible to reach mixed gas selectivity of


16

Page 17 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


about 200, propylene permeance of about 15 GPU and keep the values constant for one week. This remarkable result brought the knowledge that introduction of silver salt in inert polymers, i.e. polymers that do not have functional groups to dissolve the silver salt, is feasible; and yet, they can promote the facilitated transport of olefins. In the set of electrolyte membranes, the longest-term permeation test was performed in a PVP/AgBF4/n-octyl β-D-glucopyranoside (8G1) membrane. Along 30 days, the presence of nonionic surfactant (8G1) provided a stable membrane performance (mixture selectivity C3H6/C3H8 = 60 and mixed gas permeance of 34 GPU) with the highest permeance reported for electrolyte membranes.106 The introduction of 8G1 only avoided the Ag0 growth and agglomeration, however the reduction problems remained unsolved. Among attempts that indeed try to protect Ag+ cation against reduction, POZ/AgBF4/Al(NO3)3 membrane showed a stable performance (mixture selectivity C3H6/C3H8 = 21 and mixed gas permeance of 4.8 GPU) for 14 days.109 The highest selectivity value was found in PDMS/AgBF4 membrane with a stable performance (mixture selectivity C3H6/C3H8 = 200 and mixed gas permeance of 15 GPU) for 7 days.90 In general, the selectivity values can vary from 5 to 200 and the mixed gas permeance from 0.5 to 34 GPU.90,106,110 The time reported in long-term permeation tests ranges from 4 to 30 days.90,110,112 PVP is the most used polymer for electrolyte membranes of polar matrix; however, membranes made of PDMS, which is an inert matrix, have shown the highest selectivity values. Despite all efforts, the selectivity loss caused by Ag+ cation reduction remains unsolved for permeation tests longer than 2 weeks. Considering simultaneously selectivity, permeance, and separation stability, the best result is performed by PDMS/AgBF4 membrane, indicating that the use of inert matrixes is a promising strategy to develop new electrolyte membranes for olefin/paraffin separation. To


17


Page 18 of 59

avoid the reduction of Ag+ inside the polymer matrix, besides the silicone-based polymers, other polymer class, which is well known by its intrinsic inertness, is thought to be used as membrane material. Fluoropolymers have been used in the latest works trying to solve the problem of Ag+ instability.22,113 In general, since Ag+ is a stronger oxidant, the aim is that the polymer to be used as membrane does not have any group that could be oxidized by the Ag+ cation.

Table 1. Results of long-term permeation tests of various electrolyte membranes. Polymer

Carrier/ stabilizer Fraction (%)

Separation performance (days)

Selectivityc

Mixed gas permeance (GPU)

Olefin purity (mol%)

Reference

160

7.5

99.4

In 2001,

Electrolyte membranes of polar matrix PVP

AgBF4/DOP

4.2

50.0a*/2.0b* PVP

AgBF4/DPP

Jose al.105 4.2

135

10

99.3

50.0a*/2.0b* PVP

AgBF4/DBP

85

9

98.8

50.0a*/2.0b* PVP

AgBF4/8G1

30

50

34

98.0

AgBF4/8G1

60

34

98.4

49.9/0.2a PEP

AgNO3

16.2

5.4

50.0a

94.2

et

In 2003, Park al.106

7

et

In 2003, Park al.106

30

et

In 2001, Jose al.105

49.8/0.5a PVP

In 2001, Jose al.105

4.2

et

et

In 2006, Kang al.114

et

a - molar fraction; b - weight fraction; a* - molar fraction relates only to the polymer; b* - weight fraction relates only to the polymer; c - mixed gas (50:50 vol % of propylene/propane mixture) *Al(NO3)3·9H2O


18

Page 19 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


PVP – polyvinylpyrrolidone; PEP – poly(ethylene phthalate); POZ – poly(2-ethyl-2-oxazoline); PEO – poly(ethylene oxide); PVA – poly(vinyl alcohol); PDMS – polydimethylsiloxane DOP – dioctyl phthalate; DPP – diphenyl phthalate; DPB – dibutyl phthalate; 8G1 – n-octyl β-D-glucopyranoside.

Table 1. Continued. Polymer

Carrier/ stabilizer Fraction (%)

Separation performance (days)

Selectivityc

Mixed gas permeance (GPU)

Olefin purity (mol%)

Reference

21

4.8

95.5

In 2013,

Electrolyte membranes of polar matrix POZ

AgBF4/

14

Al(NO3)3*

Kang al.109

47.6/4.8a PEO

AgBF4/

10

10

20

90.9

Al(NO3)3*

PVP

AgCF3SO3/

4

5

0.5

83.3

Al(NO3)3*

AgCF3SO3/

4

9

0.5

90.0

Al(NO3)3*/

et

In 2015, Sung al.110

49.9/0.2a PVP

In 2015, Song al.115

49.9/0.2a

et

et

In 2016, Park and Kang112

BMImBF4 43.6/4.1/8.7a PVA

AgBF4/

6

17

11

94.0

Al(NO3)3*

In 2017, Park et al.26

49.3/1.5a Electrolyte membranes of inert matrix PDMS

AgBF4

7

200

15

73b

99.5

In 2004, Kim et al.90

a - molar fraction; b - weight fraction; a* - molar fraction relates only to the polymer; b* - weight fraction relates only to the polymer; c - mixed gas (50:50 vol % of propylene/propane mixture)


19


Page 20 of 59

*Al(NO3)3·9H2O PVP – polyvinylpyrrolidone; PEP – poly(ethylene phthalate); POZ – poly(2-ethyl-2-oxazoline); PEO – poly(ethylene oxide); PVA – poly(vinyl alcohol); PDMS – polydimethylsiloxane DOP – dioctyl phthalate; DPP – diphenyl phthalate; DPB – dibutyl phthalate; 8G1 – n-octyl β-D-glucopyranoside.

2.3. Challenges to avoid carrier instability. Taking into account the results reached at laboratory scale by using polymer membranes with silver salts45, the search for higher selectivities and permeabilities in the separation process is no more a big challenge. However, the maintenance of the membrane performance in long-term operation conditions has become a new target to be surpassed. Ag+ cations incorporated in the electrolyte membranes report instability problems related to the tendency of them to react with other species, deactivating or poisoning the agent carrier in long-term operation.28,31 The photoreduction or the exposition to reductant gases, e.g. hydrogen, inactivates the Ag+ cation due to its reduction to metallic silver (Ag0). The decontrolled formation of Ag0 in the membrane can damage it with negative influence on the separation performance. The Ag+ cation also can react with hydrogen sulfide (H2S) and acetylene (C2H2) forming undesired compounds, principally silver acetylide that is extremely explosive and can pose a significant risk to the process. The deactivation reaction of Ag+ are summarized in Table 2. It is worth emphasizing that small amounts (about 10 ppm) of these contaminants in the gas stream is enough to drastically decrease the process selectivity, in less than one week, impairing the membrane use.28

Table 2. Deactivation reaction of silver cation (Ag+). Reaction

Description uv

Ag+ + e- → Ag0

Photoreduction

2Ag+ + H2 → 2Ag0 + 2H+

Reduction by H2


20

Page 21 of 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


C2H2 + 2AgX → Ag2C2 + 2HX Formation of silver acetylide H2S + 2AgX → Ag2S + 2HX

Formation of silver sulfide

X is an anionic component of silver salt

Normally, the olefin/paraffin stream from naphtha steam cracking, which aims to be separated by membrane technology, contains some silver poisonous agents in low concentration, in ppm range. In the naphtha cracking furnace, it is necessary to operate with about 20 ppm of sulfur compounds in the feedstock to prevent the formation of undesired carbon monoxide. The function of sulfur is to passivate the nickel and iron catalysts sites in the cracking coil material of the furnace. The H2S formed in the cracked gases is removed together with the CO2 in the compression section using caustic solvents in absorption towers. Usually, CO2 and H2S concentration in the overhead stream of the absorption towers is below 0.2 ppm.6,8 Usually, hydrogen reduced compounds and acetylene (C2 and C3) are some byproducts of the SC. Hydrogen is removed at the lowest temperatures achieved in the chilling train, together with methane; they are overhead products of the demethanizer.25 The acetylene species, i.e. acetylene, methylacetylene (MA), and propadiene (PD), are removed by catalytic hydrogenation processes that transform them into more saturated hydrocarbons. The content of acetylenic compounds in the outlet of the hydrogenation process is

paraffin

Recommend Documents